Network Layer Design Issues packets can be sent.

This connection
is called a VC (virtual circuit) and the
 Store-and-Forward Packet subnet is called a virtual-circuit
Switching subnet.

A host with a packet to send  Implementation of connection

transmits it to the nearest router, Oriented Service
either on its own LAN or over a point-
to-point link to the carrier. The For connection-oriented service, we
packet is stored there until it has need a virtual-circuit subnet.
fully arrived so the checksum can be
verified. Then it is forwarded to the The idea behind virtual
next router along the path until it circuits is to avoid having to choose
reaches the destination host, where a new route for every packet sent.
it is delivered. This mechanism is Instead,when a connection is
store-and-forward packet switch-ing, established, a route from the source
as we have seen in previous machine to the destination machine
chapters. is chosen as part of the connection
setup and stored in tables inside the
 Services Provided to the routers. That route is used for all
Transport Layer traffic flowing over the connection,
1. The services should be exactly the same way that the
independent of the router teleph-one system works. When the
technology. connection is released, the virtual
2. The transport layer should be
circuit is also terminated.
shielded from the number, type,
and topology of the routers
With connection-oriented service,
present.
3. The network addresses made each packet carries an identifier
available to the transport layer telling which virtual circuit it belongs
should use a uniform numbering to.
plan, even across LANs and
WANs.  Comparison of Virtual-Circuit and
Datagram Subnets
 Implementation of
Connectionless Service

Two different organizations are

possible, depending on the type of
service offered.

If connectionless service is
offered, packets are injected into the
subnet individually and routed
independently of each other. No
advance setup is needed. In this
context, the packets are frequently
called datagrams and the subnet is
called a datagram subnet.

If connection-oriented service
is used, a path from the source
router to the destination router must Description of Routing Algorithms
be established before any data
 1 Definition: The routing algorithm
is that part of the network layer
software responsible for deciding
which output line an incoming packet
should be transmitted on.

There is enough traffic between A

and A', between B and B', and between C
and C' to saturate the horizontal links. To
maximize the total flow, the X to X' traffic
should be shut off altogether. Unfortunately,
X and X' may not see it that way. Evidently,
some compromise between global efficiency
and fairness to individual connections is
needed.
Routing decisions are made locally. When 3 Category of algorithm:
a packet arrives at router A ,it is up to A to decide nonadaptive and adaptive.
if this packet should be sent on the line to B or the
line to C. How A makes that decision is called the Nonadaptive algorithms do not
routing algorithm. Many of them exist. base their routing decisions on
measurements or estimates of the current
2.Properties of routing traffic and topology. Instead, the choice of
algorithm: correctness, simplicity, the route to use to get from I to J(for all I
robustness, stability, fairness, and and J) is computed in advance, off-line, and
optimality. downloaded to the routers when the
1) Robustness:Once a major network is booted. This procedure is
network comes on the air, it may be sometimes called static routing.
expected to run continuously for years Adaptive algorithms, in contrast,
without system-wide failures. During that change their routing decisions to reflect
period there will be hardware and software changes in the topology, and usually the
failures of all kinds. Hosts, routers, and lines traffic as well. Adaptive algorithms differ in
will fail repeatedly, and the topology will where they get their information (e.g.,
change many times. The routing algorithm locally, from adjacent routers, or from all
should be able to cope with changes in the routers), when they change the routes (e.g.,
topology and traffic without requiring all every ΔT sec, when the load changes or
jobs in all hosts to be aborted and the when the topology changes), and what
network to be rebooted every time some metric is used for optimization (e.g.,
router crashes. distance, number of hops, or estimated
2)Stability: It is also an important transit time). In the following sections we
goal for the routing algorithm. There exist will discuss a variety of routing algorithms,
routing algorithms that never converge to both static and dynamic.
equilibrium, no matter how long they run.
A stable algorithm reaches equilibrium and Routing Algorithms
stays there. • The Optimality Principle
3) Fairness and optimality may • Shortest Path Routing
sound obvious, but as it turns out, they are
• Flooding
often contradictory goals.
• Distance Vector Routing
• Link State Routing
• Hierarchical Routing
• Broadcast Routing